论文部分内容阅读
为了提高飞机利用率,航空公司在制定航班计划时会安排同一架飞机每天连续执行多个航班。当连续航班中的前序航班出现延误时,后续航班往往也会受到影响。为了研究航班计划对于航班延误波及传播的影响,构建了延误波及变化这一指标来反映延误波及增加或者减少的变动趋势。通过马尔科夫毯网络建立了航班计划的多个属性和延误波及变化的贝叶斯网络模型。利用列联表卡方检验揭示了航班计划中影响延误波及变化的主要因素为:相邻两个航段之间的计划过站时间和实际过站时间的差,前一个航段的航线类型,航段顺序。通过分析3个因素对于延误波及变化的影响,为航空公司控制延误波及和优化航班计划提出相应建议。
To improve aircraft utilization, airlines plan to schedule more than one flight per day on the same aircraft as the airline plans. When there is a delay in the pre-flight on consecutive flights, subsequent flights will often be affected. In order to study the impact of flight plan on the spread of flight delays, a measure of delay fluctuation is constructed to reflect the increasing or decreasing delay. Several attributes of the flight plan and a Bayesian network model that delays the change are established through the Markov Blanket Network. The use of contingency table chi-square test revealed that the main factors affecting the impact of delay fluctuations in flight planning are: the difference between planned transit time and actual transit time between two adjacent flight segments, the route type of the previous flight segment, Segment sequence. By analyzing the influence of three factors on the change of delay, this paper puts forward some suggestions for airlines to control delay spread and optimize flight schedule.